196 PHYSICAL GEOGRAPHY. 



there being besides some long rapids where it is not 

 necessary to unload ; and in descending to Lake Winipeg 

 the portages are about fifty, their number and even the 

 route varying with the height of water. Thousand Lakes 

 (Journ. p. 62.) is an exti*emely irregular piece of water, 

 having many extensive arms, some of which are very 

 shallow. Multitudes of islands well covered with birch, 

 aspen, arbor- vitae {Thuya occidentalism and the various 

 pine trees of the region, render the scenery pleasing. A 

 few granite knolls and mural precipices show among the 

 trees ; but many of the islets appear to be formed of sand, 

 of which sections twenty feet high occur. The lake not 

 only branches into the neighbourhood, but water com- 

 munications diverge from it in every direction, as is cus- 

 tomary in the "middle primitive belt." 



From the Thousand Lakes the canoe route keeps on 

 the border of the primitive rocks, touching on silurian 

 deposits, when it bends to the southward or westward. 

 At first it is flanked on both sides by granite. In 

 Rainy Lake there is much mica-slate ; and at its outlet 

 the stream, falling over gneiss rocks, produces the cas- 

 cade of La Chaudiere. The greater fertility of the 

 country about Nemican Lake and Rainy River show 

 the vicinity of newer formations. In the Lake of the 



barometer gave 328 feet as the height, which was a greater degree of 

 accordance between the instruments than I generally found. Major 

 Long estimates the water-shed between Lake Winipeg and Superior 

 at 1,200 feet above the tide; Major Delatield calculates the height of 

 Cold-water Lake at 505 feet, to which, if 161 be added for the Prairie 

 Portage, and 641 for Lake Superior, Ave have 1,307 feet for the height 

 of Prairie Portage above the sea. Captain Lefroj, by barometrical 

 measurements made in connection with the observatory at Toronto, 

 makes the west end of Prairie Portage 1,361 feet above the sea; but 

 the distance between the two places of observation renders the result 

 liable to some error. 



